Académique Documents
Professionnel Documents
Culture Documents
CME ARTICLE
Keywords: Background and purpose: The clinical spectrum of different neuroradiological fea-
cerebral venous thrombo- tures of cerebral sinus–venous thrombosis (CSVT) varies considerably. We sought the
sis, intracerebral hemor- relationship between different neuroradiological aspects and clinical presentations in
rhage, sinus thrombosis, these patients.
venous hemorrhagic Methods: The diagnosis of cerebral sinus–venous thrombosis has been confirmed by
infarct conventional angiography, MRI combined with MR venography following established
diagnostic criteria. We analyzed clinical data, symptoms and signs, imaging findings,
Received 14 July 2011 location and extent of the thrombus, and parenchymal lesions, retrospectively.
Accepted 31 August 2011 Results: There were 220 consecutive patients with cerebral sinus–venous thrombosis;
98 (45%) had non-lesional sinus–venous thrombosis (NL CSVT), 51 (23%) had non-
hemorrhagic infarct (NHI), 45 (20%) had hemorrhagic infarct (HI), and 26 (12%) had
intracerebral hemorrhage (ICH). In patients with hemorrhagic lesion (HI+ICH),
advanced age, headache (99%), behavioral disturbances (55%), consciousness dis-
turbances (35%), seizures (41%), and language deficits (42%) were significantly higher
than the other patients (NL+NHI) (P < 0.001). High blood pressure at admission,
puerperium, sigmoid and straight sinus thrombosis, multiple sinus and vein involve-
ment were more frequent in patients with hemorrhagic lesion than those with non-
hemorrhagic lesion. Patients with hemorrhagic lesion were more dependent or died
(32%) than the other patients (12%) (P < 0.001), and most of the patients with NL
and NHI had no disability compared with the other patients at the 3 month of follow-
up (96% and 65%; P < 0.001).
Conclusion: Headache, convulsion, behavioral disorder, seizures, and speech disor-
ders were the most frequent clinical symptoms of patients with hemorrhagic CSVT.
Specific risk factors, including pregnancy/puerberium, early and extended thrombosis
of large sinus, and presence of high blood pressure at admission, are associated with
hemorrhagic lesion and unfavorable outcome.
features in patients with CSVT in the early phase of might predispose to thrombosis, including pregnancy
sinus–venous stroke. and puerperium, oral contraceptives, malignancies,
systemic inflammatory disease, hematologic diseases,
coagulation disorders, infectious diseases, bowel dis-
Methods
ease, cardiac disease. Patients without a history of
We studied 220 consecutive patients with diagnosis of hypertension but with high blood pressure on admis-
CSVT in our Stroke Registry, who were admitted to sion (>140/85 mmHg) was recorded. Of 220 patients,
our Stroke Unit between January 1998 and May 2010, 174 (79%) were treated by anticoagulants including
retrospectively [10]. The diagnosis of CSVT was based subcutaneous non-fractionated heparin (heparin bolus
on clinical suspicion and imaging confirmation, and of 3000 U followed by continuous intravenous infusion
regarding to the clinical categories published recently to achieve an activated partial thromboplastin time
[11]. The CSVT was divided into four subgroups twice the pre-treatment value) and low-molecular
according to neuroradiological findings as following; (i) weight heparin (nadroparin dosed on the basis of body
intracerebral hematoma (ICH) was considered if the weight –180 anti-factor Xa units per kilogram daily in
patient had a confluent hemorrhagic lesion >1 cm on two divided doses) within the first 24 h of admission.
CT and MRI; (ii) venous HI was deemed if there is We did not start anticoagulation in 20% of patients
hypodensity with or without focal area of hyperdensity presenting either a major contraindication for antico-
corresponding to small hemorrhage on CT, and MRI agulation, such as major hemorrhage or clinical
fast fluid-attenuated inversion recovery (FLAIR) regression without parenchymal damage, because of the
sequence demonstrating heterogenous area of hyperin- risks and benefits of anticoagulation. The presence of
tensity and/or hypointensity, diffusion-weighted imag- clinical worsening during hospitalization was recorded,
ing (DWI) displaying predominant hyperintense area, but there was no systematic imaging data on the
and apparent diffusion coefficient (ADC) map showing occurrence of new hemorrhagic lesions after anticoag-
mixed areas of decreased and/or slightly increased ADC ulant initiation. All subjects were evaluated clinically on
values; (iii) NHI was defined if there is an ischemic 15th day, second and third months to determine the
lesion without additional lesion corresponding to patients functional final outcome. Ethic committee of
hemorrhage on initial and following repeated CT and the University Hospital approved the concept and
MRI; (iv) non-lesional CSVT (NL) was described if the design of the study (EUTF-2010).
radiological examination revealed neither an ischemic
nor hemorrhagic lesion on CT and MRI, but conven-
Statistical analysis
tional angiography or MR venography showed an
occlusion of sinus–vein system. Descriptive statistics are reported as mean and range.
All patients underwent neuroimaging examinations, Outcome analysis included short-term mortality and
including CT and MRI in the first 72 h of admission, functional outcome assessment at 15th day, 2nd and 3th
and reviewed for lesion sites, presence of venous month following discharge by the modified Rankin
infarction, and parenchymal hemorrhage. MR imaging scale (mRS): mRS(0–2), no or minor disability; mRS
sequences were performed with a 1.5-T MR unit (3–6), dependent and death; mRS (6), death. One-way
2
(Magnetom Vision; Siemens Medical systems, Erlan- ANOVA or Pearson v test when appropriate were used
gen, Germany) including T1- and T-2 weighted to assess relationships between variables. The post hoc
sequences, FLAIR imaging sequence, and DWI se- analysis was made by Bonferroni and Tukeys tests and
quences with ADC calculations. MRA and magnetic pairwise multiple comparisons to assess which means
resonance venography (three-dimensional time-of-flight differ. A second analysis with Pearson v2 test and 95%
sequences) were obtained during the same imaging CI was made by separating patients into two groups;
session. Two senior neuroradiologists were blinded to patients without hemorrhagic lesion (NL +NHI) and
the clinical information. Interobserver agreement for those with hemorrhagic infarct (HI+ICH), to estimate
the radiological assessment according to criteria men- risk factors in the way cause of hemorrhage. Logistic
tioned previously was higher between neuroradiologists regression analysis was conducted to select the best
(j = 0.78). working correlation structure for the unstandardized
We recorded demographic and clinical data, lesions and standardized coefficients (B) and 95% CI between
at admission on computed tomographic and MRI two groups. In the regression analysis, different models
scans, (left, right or bilateral hemisphere, single or were constructed by probability-of-F-to-enter £0.05 to
multiple), location of thrombosis, consciousness alter- select the best working correlation between groups by
ations, clinical symptoms, seizures, and visual loss, any entering variables such as age, sex, high blood pressure,
associated illness and medication or condition that pregnacy/puerberium, sigmoid and straight sinus
thrombosis, multiple sinus and sinus–veins involve- with NHI and NL (22% and 15% vs. 4% each;
ment. All data were analyzed with SPSS 17.0 (SPSS P = 0.002). Headache, sensorial deficits, seizures, dys-
Inc., Chicago, IL, USA) statistical software. arthria and aphasia, and behavioral deficits (anxiety,
depression) were significantly higher in patients with
hemorrhagic lesion (Table 1). Consciousness loss was
Results
significantly higher in patients with HI and ICH com-
Among 220 patients, 152 (69%) patients were female, pared those with NL CSVT and NHI (20% and 27%
and the mean age was 44 ± 7 years old (range, 18– vs. 0% and 2%; P = 0.001). Sigmoid sinus involve-
75 years). Of these, 121 (55%) patients had only sinus ment was significantly more frequent in patients with
thrombosis, 77 (35%) had concomitant sinus and vein NHI and HI (53% and 64%; P = 0.001). The fre-
thrombosis, and 22 (10%) had only vein thrombosis. quency of multiple sinus involvement was more higher
Ninety-eight patients (45%) had no lesion on CT/MRI, in patients with NHI, HI, and ICH than those with NL
51 (23%) had NHI, 45 (21%) had HI, and 26 (12%) CSVT (73% and 62% each vs. 21%; P = 0.001). The
presented ICH. involvement of straight sinus was significantly higher in
High blood pressure at admission was found more patients with ICH and NHI (31% and 29%;
frequently in patients with HI and ICH compared those P = 0.001). (Table 2). Post hoc range and pairwise
Table 1 Demographic features and clinical findings in patients with cerebral sinus–vein thrombosis regarding to non-lesional venous stroke, non-
hemorrhagic infarct, hemorrhagic infact, and intracerebral hemorrhage
NL, non-lesional sinus–venous thrombosis; NHI, non-hemorrhagic infarct; HI, hemorrhagic infarct; ICH, intracerebral hemorrhage; mRS
modified Rankin score.
Values in parenthese are percentage of column.
a
Pearson v2 test; df:3.
b
Significantly different group mean at Post hoc multiple comparison tests.
c
At the 3 month of follow-up.
Table 2 Etiology and sinus–venous involvement regarding to non-lesional venous stroke, non-hemorrhagic infarct, hemorrhagic infact, and
intracerebral hemorrhage
Etiology
Pregnancy/puerperium 7 (7) 5 (10) 15 (33)a 7 (27) 0.001b
Oral contraceptives 11 (11) 4 (8) 7 (16) 6 (23) 0.2
SLE 12 (12) 4 (8) 6 (13) 3 (12) 0.8
APA 8 (8) 2 (4) 4 (9) 2 (8) 0.8
Hematologic disorders
Protein-C deficiency 6 (6) 3 (6) 3 (7) 0 0.6
Protein-S deficiency 9 (9) 1 (2) 5 (11) 2 (8) 0.3
Factor V mutation 9 (3) 2 (2) 1 (2) 1 (4) 0.3
Antithrombin III deficiency 3 (3) 0 4 (9) 1 (4) 0.1
Prothrombin II mutation 3 (1) 1 (2) 1 (2) 1 (4) 0.9
Polycythemia 0 4 (8) 3 (7) 0 0.02
MTHFR 677T mutation 3 (3) 3 (6) 5 (11) 0 0.1
Dehydratation 0 4 (8) 1 (2) 2 (8) 0.03
Behçets disease 9 (9) 5 (10) 1 (2) 3 (12) 0.4
Intracranial mass 5 (5) 1 (2) 1 (2) 0 0.5
Meningioma 3 (3) 1 (2) 1 (2) 0
Astrositoma 2 (2) 0 0 0
Sinus involved (n = 121)c
Superior sagittal sinus 40 (41) 21 (41) 23 (51) 13 (50) 0.6
Inferior sagittal sinus 12 (12) 14 (28) 8 (18) 5 (19) 0.1
Transverse sinus 48 (49) 28 (55) 23 (51) 14 (54) 0.9
Sigmoid sinus 27 (27) 27 (53) 29 (64)a 11 (42) 0.001
Juguler sinus 10 (10) 7 (14) 6 (13) 1 (4) 0.6
Straight sinus 5 (5) 15 (29) 9 (20) 8 (31)a 0.001
Cavernous sinus 9 (9) 2 (4) 1 (2) 2 (7) 0.4
Multiple sinus involvement 21 (21) 37 (73) 28 (62) 16 (62) 0.001
Veins involved (n = 22)c 15 (15) 1 (2) 5 (11) 1 (4) 0.05
Cortical venous system 8 (8) 0 2 (4) 1 (4)
Deep venous system 7 (2) 1 (2) 3 (7) 0
Combined vein+sinus (n = 77)c 12 (12) 25 (49) 21 (47) 19 (73)a 0.001
multiple comparisons tests determined that female patients with hemorrhagic lesions (Table 3). Logistic
gender was significantly higher in patients with HI, and regression analysis of risk factors showed a strong
the mean differences in the other clinical and risk fac- correlation between age, sex, high blood pressure at
tors were because of the means of patients with HI and admission, sinus–vein thrombosis, multiple sinus
ICH. involvement, and hemorrhagic lesions (Table 4).
Parietal lobe involvement was present in 39% (20/51) Most of the patients with NL CSVT and NHI pre-
of patients with NHI, in 44% of those with HI, and in sented complete and partial recovery (96% and 65%),
58% of those with ICH. Frontal lobe was affected in while patients with HI and ICH had more severe dis-
23% of patients with NHI, in 31% of those with HI, ability or death (mRS 3–6) (42% and 35%; P = 0.001)
and in 27% of those with ICH. Temporo-occipital lobes at the 3 month of follow-up. Eight of 174 (4.6%)
were involved in 35% of patients with NHI, in 13% of patients who were treated by anticoagulants and 2 of 46
those with HI, and in 11% of those with ICH. (4.3%) patients without anticoagulant treatment died
Comparison of patients with hemorrhagic lesion and (OR = 1.1, 95% CI 0.2–5.2; P = 0.7). Among 71
without showed that age, female gender, high pressure patients with hemorrhagic lesions, 14% (8/58) of pa-
on admission, pregnancy/puerperium, sigmoid sinus tients, who were treated by anticoagulants and 15% (2/
involvement, multiple sinus thrombosis, and concomi- 13) of patients who were treated by conservative
tant sinus–veins thrombosis were significantly higher in treatment, died (OR = 0.9, 95% CI 0.2–4.7; P = 0.6).
Table 3 Risk and vascular factors in patients with hemorrhagic lesion vs. those without hemorrhagic lesion
tion had an adverse effect on the clinical outcome in 2. Bousser MG. Cerebral venous thrombosis: diagnosis and
patients with hemorrhagic brain lesions cannot be management. J Neurol 2000; 247: 252–258.
3. De Bruijn SFTM, de Haan RJ, Stam J, for the Cerebral
derived from the presented data. Moreover, hemor-
Venous Sinus Thrombosis Study Group. Clinical features
rhage itself is an independent predictor for a poor and prognostic factors of cerebral venous and sinus
clinical outcome. Limited data from randomized con- thrombosis in a prospective series of 59 patients. J Neurol
trolled clinical trials in combination with observational Neurosurg Psychiatry 2001; 70: 105–108.
data on outcomes and bleeding complications of anti- 4. Coutinho JM, Ferro JM, Canhão P, et al. Cerebral ve-
nous and sinus thrombosis in women. Stroke 2009; 40:
coagulation support a role for anticoagulation in the
2356–2361.
treatment of CSVT [15–21]. It is logical that in the 5. Ferro JM, Correia M, Rosas MJ, Pinto AN, Neves G,
special situation of a patient with a recent hemorrhagic Cerebral Veno Thrombosis Portuguese Colloborative
lesion, the clinician must balance the risks and benefits Study Group (Venoport). Seizures in cerebral vein and
of anticoagulation, depending on the clinical situation. dural sinus thrombosis. Cerebrovasc Dis 2003; 15: 78–83.
6. Preter M, Tzourio C, Ameri A, Bousser MG. Long term
In the clinical routine practice, low-intensity anticoag-
prognosis in cerebral venous thrombosis. Follow-up of 77
ulation may be considered if possible in favor of no patients. Stroke 1996; 27: 243–246.
anticoagulation until such time as it might be safe to use 7. Biousse V, Tong F, Newman NJ. Cerebral venous
full-intensity anticoagulation. thrombosis. Curr Treat Options Neurol 2003; 5: 409–420.
The outcome of patients with NL, NHI was good, 8. Théaudin M, Crassard I, Bresson D, et al. Should
decompressive surgery be performed in malignant cerebral
and more than two-thirds was discharged with complete
venous thrombosis? A series of 12 patients. Stroke 2010;
or partial recovery, while half of the patients with HI 41: 727–731.
was independent at the 3rd month of follow-up. 9. Ferro JM, Canhao P, Stam J, Bousser MG, Barinagarr-
Approximately, two-thirds of the patients having ICH ementeria F, ISCVT Investigators. Prognosis of cerebral
were dependent, which could be a predictor of an vein and dural sinus thrombosis: results of the Interna-
tional Study on Cerebral Vein and Dural Sinus Throm-
unfavorable outcome in these patients. The ISCVT trial
bosis (ISCVT). Stroke 2004; 35: 664–670.
[9,22] reported that 3.4% of patients died within 10. Kumral E, Ozkaya B, Sagduyu A, Sirin H, Vardarli E,
30 days from symptom onset, and risk factors were Pehlivan M. The Ege Stroke Registry: a hospital-based
depressed consciousness, altered mental status, tran- study in the Aegean Region, Izmir, Turkey. Cerebrovasc
stentorial herniation secondary to a large hemorrhagic Dis 1998; 8: 278–288.
11. Saposnik G, Barinagarrementeria F, Brown RD, et al.
lesion, and diffuse brain edema, supporting the findings
Diagnosis and management of cerebral venous thrombo-
found in our series. Another recent study showed that sis: a statement for healthcare professionals from the
patients with non-hemorrhagic lesion had a better American Heart Association/American Stroke Associa-
prognosis in the acute phase and at the end of follow-up tion. Stroke 2011; 42: 1158–1192.
than those with hemorrhagic lesion [22]. 12. Frerichs KU, Deckert M, Kempski O, Schurer L, Ein-
haupl K, Baethmann A. Cerebral sinus and venous
In conclusion, risk factors including involvement of
thrombosis in rats induces long-term deficits in brain
large and multiple sinus, pregnancy/puerperium, and function and morphology: evidence for a cytotoxic gene-
advanced age may predispose patients to sinus–venous sis. J Cereb Blood Flow Metab 1994; 14: 289–300.
hemorrhagic stroke and unfavorable outcome. Patients 13. Weih M, Junge-Hulsing J, Mahraein S, Ziemer S, Ein-
admitted for CSVT having hemorrhagic lesion should haupl KM. Hereditary thrombophilias in ischemic stroke
and sinus vein thrombosis: diagnostic, therapy and meta
be considered as being at risk and deserve additional
analysis. Nervenarzt 2000; 71: 936–945.
close monitoring. New treatment options, besides 14. Breteau G, Mounier-Vehier F, Godefroy O, et al. Cere-
anticoagulation, should be considered cautiously bral venous thrombosis 3-year clinical outcome in 55
because it is necessary to identify predictors of hemor- consecutive patients. J Neurol 2003; 250: 29–35.
rhagic lesions to avoid from the death and poor 15. Einhäupl K, Stam J, Bousser M-G, et al. EFNS guideline
on the treatment of cerebral venous and sinus thrombosis
outcome.
in adult patients. Eur J Neurol 2010; 17: 1225–1313.
16. Wingerchuk DM, Wijdicks EF, Fulgham JR. Cerebral
venous thrombosis complicated by hemorrhagic infarc-
Disclosure of conflict of interest
tion: factors affecting the initiation and safety of antico-
The authors declare no financial or other conflict of agulation. Cerebrovasc Dis 1998; 8: 25–30.
17. Fink JN, McAuley DL. Safety of anticoagulation for
interests.
cerebral venous thrombosis associated with intracerebral
hematoma. Neurology 2001; 57: 1138–1139.
18. Frey JL, Muro GJ, McDougall CG, Dean BL, Jahnke
References
HK. Cerebral venous thrombosis: combined intra-
1. Bousser MG, Chiras J, Bories J, Castaigne P. Cerebral thrombus rtPA and intravenous heparin. Stroke 1999; 30:
venous thrombosis-a review of 38 cases. Stroke 1985; 16: 489–494.
199–213.
19. Girot M, José M, Ferro JM, et al. Predictors of outcome 21. Ferro JM, Canhão P, Stam J, et al. Delay in the diagnosis
in patients with cerebral venous thrombosis and intrace- of cerebral vein and dural sinus thrombosis: influence on
rebral hemorrhage. Stroke 2007; 38: 337. outcome. Stroke 2009; 40: 3133–3138.
20. Canhão P, Ferro JM, Lindgren AG, et al. Causes and 22. Ferro JM, Canhão P, Bousser M-G, et al. Cerebral venous
predictors of death in cerebral venous thrombosis. Stroke thrombosis with nonhemorrhagic lesions: clinical corre-
2005; 36: 1720–1725. lates and prognosis. Cerebrovasc Dis 2010; 29: 440–445.